8000 Commits · evdevdev/tensorflow · GitHub

Commits

Commits on Jan 23, 2016

Merge pull request #855 from zffchen78/master.
Vijay Vasudevan
committed
Avoids an extra copy of the graph def.
Zhifeng Chen
committed
Merge pull request #854 from eguven/patch-1.
Vijay Vasudevan
committed
fix typos in docstrings

eguven
authored and
Vijay Vasudevan
committed
Merge pull request #849 from wasserfuhr/patch-2.
Vijay Vasudevan
committed
Update ops.py

wasserfuhr
authored and
Vijay Vasudevan
committed
Merge pull request #844 from wasserfuhr/patch-1.
Vijay Vasudevan
committed
Update framework.md

wasserfuhr
authored and
Vijay Vasudevan
committed
Merge pull request #847 from benoitsteiner/master.
Vijay Vasudevan
committed
Ensure the code compiles when AVX2 instructions are enabled.
benoitsteiner
committed
Merge pull request #846 from benoitsteiner/master.
Vijay Vasudevan
committed
Improved the performance of spatial convolutions on CPU:
benoitsteiner
committed
Merge commit for internal changes
Vijay Vasudevan
committed
Use an iterator to walk down the set of free chunks, so that deleting

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed
Fixes for OSS build.
Vijay Vasudevan
committed
Pin constant holding data to CPU. This makes GPU version 50-100x faster per step. Fixes #837

yaroslavvb
authored and
Vijay Vasudevan
committed
Implement a promise-based RequestManager class, will be the workhorse of a new TensorFlow backend.

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed
Disentangle the GPU code from the CPU code. This means a few things:

josh11b
authored and
Vijay Vasudevan
committed
Improved the performance of tensor contractions and spatial convolutions on CPU:

benoitsteiner
authored and
Vijay Vasudevan
committed
Added a script for rough estimation of the computation cost and parameter size of a CNN.

petewarden
authored and
Vijay Vasudevan
committed
Add symbolic gradient function for Softmax.

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed
Fixes a bug in Queue cancellation.

mrry
authored and
Vijay Vasudevan
committed

Commits on Jan 22, 2016

Merge commit for internal changes
Vijay Vasudevan
committed
Rollback of "Adding histogram_ops module containing one Op: histogram_fixed_width, which updates a histogram Variable with new_values."
Vijay Vasudevan
committed
Change uses of TensorShape::ShortDebugString to DebugString

girving
authored and
Vijay Vasudevan
committed
Make cuda_dnn work with R4.

zheng-xq
authored and
keveman
committed
Add TensorArray data_flow_ops wrapper and TensorArray gradients.

ebrevdo
authored and
Vijay Vasudevan
committed
Add TODO to fix the situation where core_cpu_internal includes

josh11b
authored and
Vijay Vasudevan
committed
Adding histogram_ops module containing one Op: histogram_fixed_width, which updates a histogram Variable with new_values.

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed
Move #include <vector> out of port.h to users of std::vector<>.

josh11b
authored and
Vijay Vasudevan
committed
Export ToGraphDef() from function.cc

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed
Need to check if the op is in a loop context.

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed
Fix a test for Python 3

girving
authored and
Vijay Vasudevan
committed
Drop unused gemmlowp dependency.

josh11b
authored and
Vijay Vasudevan
committed
Fix and re-enable tests for matrix_triangular_solve op.

A. Unique TensorFlower
authored and
Vijay Vasudevan
committed

0